NEUROSURGICAL FOCUS Neurosurg Focus 48 (2):E8, 2020

EDITORIAL Probing the tract organization of language: Heschl’s gyrus fiber intersection area

Martina L. Mustroph, MD, PhD,1 Leo R. Zekelman, BA,2 and Alexandra J. Golby, MD1,3

Departments of 1Neurosurgery and 3Radiology, Brigham and Women’s Hospital, Harvard Medical School; and 2Speech and Hearing Bioscience and Technology, Division of Medical Sciences, Harvard Medical School, Boston, Massachusetts

n 1878, Richard Ladislas Heschl was the first author pass through a high-density tract area, which they term to describe the anterior transverse temporal gyrus, in the “Heschl’s gyrus fiber intersection area (HGFIA).” Über die vordere quere Schläfenwindung des men- These tracts are the anterior arcuate fasciculus, the middle schlichenI Grosshirns.7 This gyrus, known today as Hes- longitudinal fasciculus, the acoustic radiation and U fi- chl’s gyrus, is understood to contain the primary auditory bers connecting Heschl’s gyrus with the middle temporal cortex, as it reliably exhibits tonotopic maps3 (i.e., how gyrus and, beneath this, the inferior fronto-occipital fas- sound frequencies map to different locations) presumably ciculus and the optic radiation. The article contains high- originating from the tonotopic maps encoded by the co- quality cerebral hemisphere dissection panels which show chleae. Myriad investigations of Heschl’s gyrus have been these tracts. carried out related to many aspects of auditory process- The authors relate their anatomical findings of fiber re- ing, including developmental maturation,15 normal audi- lationships to language theory. Specifically, they note that tory processing,10 disorders of auditory processing such their work supports the hypothesis that the primary audi- as autism,18 pitch processing in musicians,20 and auditory tory area directly contacts the anterior arcuate fasciculus hallucinations in schizophrenia,2 among many others. A posteriorly. This contact between the primary auditory number of techniques have been used to examine the au- area and the anterior arcuate fasciculus could be viewed ditory cortex, including transcranial direct current stimu- as support for the existence of the external and internal lation,11 PET,13 magnetoencephalography,9 and functional loops of auditory feedback and the articulatory loop es- MRI (fMRI).16 Techniques used to analyze white matter sential for speech articulation. Hickok and Poeppel pro- within and in contact with the auditory cortex, such as posed a framework of dorsal and ventral streams involved volumetric morphometry,17 electron microscopy,1 and dif- in language production,8 which itself rests on Liberman’s fusion MRI,6 often reveal an asymmetry between the 2 motor theory of speech perception.14 Inspired by emerg- hemispheres, which is hypothesized to be related to lan- ing understanding of visual processing via distinct ven- guage lateralization. tral and dorsal streams, Hickock and Poeppel proposed 2 In this article, Fernández, Velásquez, García Porrero, processing streams for language: a ventral stream, which Marco de Lucas, and Martino present converging evi- they postulated is involved in mapping sound onto mean- dence from cadaveric fiber dissections, in vivo diffusion ing (semantic processing), and a dorsal stream, involved MRI tractography, and intraoperative subcortical stimula- in mapping sound onto articulatory-based representations tion to describe the complex white matter architecture in (phonemic processing). In the present article, Fernández et the posterior region of the superior temporal lobe around al. offer the conjecture that the HGFIA is related to these Heschl’s gyrus and the acoustic radiations.5 The authors 2 streams of language; superficially, to the dorsal phono- performed detailed dissections in 8 cadaveric hemispheres logical stream via the long and anterior segments of the (4 left and 4 right). They also acquired and analyzed data arcuate fasciculus, and deeply, to the ventral stream via from diffusion MRI probabilistic tractography in 8 con- the inferior fronto-occipital fasciculus. Additionally, the trol hemispheres from 4 healthy subjects and intraopera- authors note that the U fibers connecting the superior and tive electrical stimulation mapping in 6 patients with left middle temporal gyri may represent a direct connection temporal lobe gliomas who underwent awake resections. between the primary auditory area and arcuate fasciculus, Based on these data, the authors propose that 5 fiber tracts which could explain auditory information propagation to

ACCOMPANYING ARTICLE DOI: 10.3171/2019.11.FOCUS19778.

©AANS 2020, except where prohibited by US copyright law Neurosurg Focus Volume 48 • February 2020 1 Editorial the middle temporal gyrus and from there via the long superior temporal lobe, the HGFIA as described by the segment of the arcuate fasciculus to the ventral premotor authors. In showing that resection of the superior temporal cortex and the inferior frontal gyrus. gyrus anterior to Heschl’s gyrus did not result in language The work presented in this article raises additional deficits despite resection of a considerable portion of mid- questions regarding the role of Heschl’s gyrus in lan- dle longitudinal fasciculus, the work supports the hypoth- guage function. The demonstration that stimulation of the esis that the middle longitudinal fasciculus is not essential HGFIA at cortical and subcortical areas results in anomia for language processing. Thus, the work has important and semantic paraphasias, in addition to the detailed ana- practical implications for approaches to tumor resections tomical descriptions of these tracts, prepares the ground in the dominant temporal lobe. Future work remains to for investigations regarding the neural architecture of further elucidate the exact function of each of the 5 tracts language supported by these 5 tracts. The authors note that comprise the area, including differences in dominant previous work in which stimulating the long segment of and nondominant hemispheres. For example, diffusion the arcuate fasciculus caused phonemic paraphasias.4,12,19 imaging with high spatial and angular resolution, perhaps With this work established, it would be interesting to see in cadaveric specimens in which extended image acquisi- investigations into the specific language disturbances that tions can be obtained, could build on the proposed fiber result from stimulation of each of these 5 tracts individu- organization. We eagerly look forward to seeing what fu- ally. For example, future studies could investigate whether ture work this article inspires. stimulation of the anterior segment of the arcuate fascicu- https://thejns.org/doi/abs/10.3171/2019.11.FOCUS19886 lus results in speech comprehension and articulation defi- cits associated with auditory feedback, as the authors de- References scribe. Is it possible to separately stimulate each of these 5 1. Anderson B, Southern BD, Powers RE: Anatomic asymme- tracts? And would such stimulation result in phonological, tries of the posterior superior temporal lobes: a postmortem morphological, syntactic, semantic, or pragmatic language study. Neuropsychiatry Neuropsychol Behav Neurol comprehension and production deficits? In future awake 12:257–254, 1999 surgeries attempts could be made to use neuronavigation 2. Chen X, Liang S, Pu W, Song Y, Mwansisya TE, Yang Q, and intraoperative imaging to locate and stimulate each et al: Reduced cortical thickness in right Heschl’s gyrus tract as precisely and separately as possible while asking associated with auditory verbal hallucinations severity in first-episode schizophrenia. BMC Psychiatry 15:152, 2015 patients to perform strategic tasks designed to elicit asso- 3. Da Costa S, van der Zwaag W, Marques JP, Frackowiak RS, ciated language deficits. Clarke S, Saenz M: Human primary auditory cortex follows The authors acknowledge that the sample sizes in the the shape of Heschl’s gyrus. J Neurosci 31:14067–14075, present work (8 cadaver hemispheres, 8 healthy control 2011 hemispheres, and 6 left temporal glioma patients who un- 4. Duffau H, Moritz-Gasser S, Mandonnet E: A re-examina- derwent intraoperative stimulation testing) were limited. tion of neural basis of language processing: proposal of a As a result, the reported fiber tract findings are not broken dynamic hodotopical model from data provided by brain down by sex or language proficiency, including effects on stimulation mapping during picture naming. Brain Lang 131:1–10, 2014 specific languages spoken by multilingual patients under- 5. Fernández L, Velásquez C, García Porrero JA, de Lucas going awake surgery. While the authors clearly state the EM, Martino J: Heschl’s gyrus fiber intersection area: a new handedness of the surgical tumor patients, handedness of insight on the connectivity of the auditory-language hub. healthy subjects in the study is not specified; in fact, cere- Neurosurg Focus 48(2):E7, 2020 bral hemisphere is a between-subject variable only among 6. Glasser MF, Rilling JK: DTI tractography of the human the 8 dissected human cadaver specimens, preventing brain’s language pathways. Cereb Cortex 18:2471–2482, comparisons of tract size or connections by hemisphere 2008 7. Heschl RL. Über die vordere quere Schläfenwindung des or handedness. Notably, no formal between-hemisphere menschlichen Grosshirns. Vienna: Braunmüller, 1878 analysis was conducted in the healthy control subjects who 8. Hickok G, Poeppel D: Dorsal and ventral streams: a frame- underwent DTI. Future work could establish the locations work for understanding aspects of the functional anatomy of of the functional language cortex in the healthy subjects language. Cognition 92:67–99, 2004 and patients by using fMRI and show how these locations 9. Houde JF, Nagarajan SS, Sekihara K, Merzenich MM: relate to the fiber tracts. Correlating fMRI and language Modulation of the auditory cortex during speech: an MEG lateralization with tract findings—particularly the ques- study. J Cogn Neurosci 14:1125–1138, 2002 10. Khalighinejad B, Herrero JL, Mehta AD, Mesgarani N: tion of whether the HGFIA differs when language is co- Adaptation of the human auditory cortex to changing back- dominant compared to when it is lateralized—would be an ground noise. Nat Commun 10:2509, 2019 interesting future investigation, as would comparing the 11. Kunzelmann K, Meier L, Grieder M, Morishima Y, Dierks tracts across a wider age range (the age of cadaver speci- T: No effect of transcranial direct current stimulation of mens is expectedly narrow, 63–82 years; no age range is the auditory cortex on auditory-evoked potentials. Front specified for the healthy controls or brain tumor patients). Neurosci 12:880, 2018 Taken together, the article presents detailed white 12. Leclercq D, Duffau H, Delmaire C, Capelle L, Gatignol P, matter fiber descriptions based on anatomic dissections, Ducros M, et al: Comparison of diffusion tensor imaging tractography of language tracts and intraoperative subcorti- diffusion MRI tractography, and functional subcortical cal stimulations. J Neurosurg 112:503–511, 2010 mapping. Understanding of the numerous tracts in this 13. Lee JS, Lee DS, Oh SH, Kim CS, Kim JW, Hwang CH, et al: region could contribute an interesting perspective on the PET evidence of neuroplasticity in adult auditory cortex of language function of a specialized area of the posterior- postlingual deafness. J Nucl Med 44:1435–1439, 2003

2 Neurosurg Focus Volume 48 • February 2020